Red 40

Red 40 is the most used synthetic colour in the American food supply. It is also the one where two regulators looked at broadly the same evidence and did two different things, which makes it a better page than the usual food colour entry.

The part almost nobody knows

Red 40 is a certified colour. That word is doing specific work.

Certified colours are listed in a part of the regulations separate from the exempt ones, and the difference is not cosmetic. A batch of a certified colour cannot be sold until a sample has been submitted to FDA and tested against a published specification. The specification for Red 40 sets limits on volatile matter, water-insoluble matter, higher and lower sulfonated subsidiary colours, named reaction by-products, and heavy metals.

So every batch is examined by the regulator before it reaches a factory. Very little else in food is controlled that way. Colours exempt from certification, which includes most of the ones marketed as natural, are not.

That is not an argument that Red 40 is preferable. It is an argument that “synthetic colour” and “unexamined” are not the same thing, and that the intuition running the other way is worth checking.

Where the two regulators part company

The Southampton study, published in 2007, gave about 300 children drinks containing a mix of colours and a preservative and measured behaviour. It is the study behind everything that followed.

The European Union responded by requiring the six colours involved to carry a statement, or be eliminated: the warning is that the colour “may have an adverse effect on activity and attention in children”.

FDA convened its Food Advisory Committee on 30 and 31 March 2011 to consider the same question. The committee “concluded that the evidence was too inconclusive to link food colors to hyperactivity and too insufficient to recommend warning labels for products containing artificial food colors”.

Two conclusions, one body of evidence. It is worth reading the source here carefully: the account above is from a commentary by a researcher who testified before that committee arguing for the opposite conclusion, and who says so in the text. He also makes the point that the question FDA framed was whether colours cause hyperactivity generally, rather than whether a subgroup of children responds, and that those are different questions with different answers.

What that leaves a parent with

Not a verdict, which is uncomfortable but honest.

The finding is about a subgroup rather than about children generally, the effect sizes are modest, and the two regulators genuinely disagree about whether that is enough to put on a label. Anyone telling you this is settled in either direction is ahead of the evidence.

What is settled is that removing it is easy and costs nothing but colour. If a child of yours seems to react, an elimination trial is a reasonable thing to run and it is the kind of experiment a parent can actually do, which is worth more here than a general statement.

The names and the neighbours

On US labels it is FD&C Red No. 40 or Red 40. In Europe it is E129 or Allura Red AC. In cosmetics it may be CI 16035. The lake form is the same colour bound to a substrate so it does not dissolve, used in coatings and icings.

Tartrazine is the yellow member of the same six and its file is similar. Red 3 is a different case entirely: its US listing was revoked in 2025 under a statutory rule about animal carcinogenicity, which is a legal mechanism rather than a behavioural finding, and the distinction is set out on the banned-elsewhere page.